Version française
Home     About     Download     Resources     Contact us    

This site is updated infrequently. For up-to-date information, please visit the new OCaml website at

Browse thread
[Caml-list] Doubly-linked list
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: 2002-08-21 (08:07)
From: Diego Olivier Fernandez Pons <Diego-Olivier.FERNANDEZ-PONS@c...>
Subject: Re: Polymorphic recursion 9Was Re: [Caml-list] Doubly-linked list)
Brian Rogoff a écrit

> Oh I'm not arguing that point, I totally agree that it's omission is a 
> bad thing. However, not everyone agrees, since you it becomes a lot tougher
> to build a monomorphizing compiler if you allow it, though it has been 
> suggested that the same tricks you use to manually remove polymorphic
> recursion could be used in an SSC (sufficiently smart compiler). 

I do not agree with your analysis since I really do not believe anyone
could think that polymorphic recursion is useless. But it is a
_difficult_ subject and the Caml Team is working on it (you can read
their research summary)

Actually, I do not even think that including some tricks in the
compiler is a manageable solution.

> Anyways, since OCaml 3.05 allows first class polymorphism on records, you
> don't even need to use "polymorphic recots" to get non-uniform recursion; 
> simply mapping the OOP to records does the trick. Here's the first example
> from OKasaki which uses polymorphic recursion, done in OCaml with this 
> direct approach 


Your trick is very interesting but I am afraid I cannot include that
kind of code in a library whose purpose is also to be a pedagogical
support for a data structure course in Caml and an example for
beginners (maybe released as independent modules ?)

        Diego Olivier

To unsubscribe, mail Archives:
Bug reports: FAQ:
Beginner's list: